User:Frostburn/Collatz scales: Difference between revisions

Frostburn (talk | contribs)
Add notes about the 3/1 iteration being special and give a counter example for 5/3 iteration.
Frostburn (talk | contribs)
→Additional results: Make additional conjectures.
Line 68: Line 68:


== Additional results ==
== Additional results ==
Variations of the recursion do not seem work for all seeds e.g.
Not all variations of the recursion work for all seeds e.g.


<math>
<math>
Line 90: Line 90:
</pre>
</pre>
But produces an infinite scale for <math>\frac{19}{17}</math>.
But produces an infinite scale for <math>\frac{19}{17}</math>.
=== Additional conjectures ===
The recursion:
<math>
s_{n+1} = \frac{p_{n+1}}{q_{n+1}} = \frac{1 + 3 p_n}{4 q_n}\ \mathrm{red}\ 5
</math>
produces finite scales:
<math>\left\{ s_n | n \in \mathbb{N} \right\} \cup \left\{ 5 \right\}</math>
for all starting values <math>s_0</math>.
Evidence
<pre>
! varcollatz1_3_4_5_seed_4_3.scl
!
Collatz scale 1, 3, 4, 5 variation of 7/5
13
!
17/16
11/10
11/8
7/5
7/4
2/1
37/16
5/2
49/16
13/4
65/16
17/4
5/1
</pre>
The recursion:
<math>
s_{n+1} = \frac{p_{n+1}}{q_{n+1}} = \frac{1 + 5 p_n}{4 q_n}\ \mathrm{red}\ 3
</math>
produces finite scales:
<math>\left\{ s_n | n \in \mathbb{N} \right\} \cup \left\{ 3 \right\}</math>
for all starting values <math>s_0</math>.
Evidence
<pre>
! varcollatz1_5_4_3_seed_12_7.scl
!
Collatz scale 1, 5, 4, 3 variation of 12/7
13
!
383/336
7/6
479/336
3/2
12/7
599/336
2/1
61/28
107/48
153/56
11/4
67/24
3/1
</pre>
There are many (too many to list) other variations of the recursion that seem to work too, but these stood out.